Primary endpoint

Definition

A primary endpoint is the main outcome measure that a clinical trial or research study is designed to evaluate. The primary endpoint is specified in the study protocol (study plan) before the study begins; changing the primary endpoint after the study has commenced or after results are known constitutes scientific misconduct. Primary endpoints are objective and measurable (e.g., change in blood glucose levels, change in body weight, time to disease progression) rather than subjective (e.g., 'feeling better'). For a peptide study, the primary endpoint might be 'percentage change in glycaemic control' or 'percentage weight loss' or 'time to disease remission,' depending on the indication. The sample size of the study is calculated based on the expected effect size for the primary endpoint; if the study is too small, it may fail to detect a true effect on the primary endpoint (Type II error). Regulatory approval is typically based on whether the treatment produces a statistically significant effect on the primary endpoint. Secondary endpoints are additional outcome measures that provide supporting evidence but are not required for approval.

Specifying the primary endpoint in advance prevents researchers from data-dredging (analysing many outcomes, reporting only those that are positive, and claiming the positive ones were primary objectives). Regulatory agencies such as the TGA and FDA require that the primary endpoint be pre-specified and justified. If a trial fails to show benefit on the primary endpoint, it is considered a failed trial, even if secondary endpoints show promising results. This conservative approach protects patients from ineffective treatments.

Primary endpoints must be clinically relevant and directly reflect the intended therapeutic benefit. For example, for a diabetes peptide, the primary endpoint might be HbA1c (three-month average blood glucose) or change in body weight, both directly related to metabolic function. Surrogate endpoints (biomarkers believed to predict clinical benefit but not directly demonstrating benefit) are sometimes accepted as primary endpoints, but they require strong evidence that the surrogate correlates with true clinical improvement.